A strength Pareto Evolutionary Algorithm (SPEA) for multi-mission radar waveform optimization

This paper furthers the development of Evolutionary Computation, specifically Genetic Algorithms (GA's) applied to the waveform design of simultaneously transmitted orthogonal waveforms. The determination of a suite of “optimal” waveforms in the Pareto sense is found a priori for a single platform radar system performing multiple radar missions simultaneously. The waveform suite is determined by utilizing a Strength Pareto Evolutionary Algorithm to find solutions to a Multi-Objective Optimization Problem (MOP). The objectives to be optimized are dictated by the particular missions of interest. The mapping of these objective functions to actual radar performance parameters is used in the SPEA to determine how best to simultaneously perform multiple radar missions such as GMTI, AMTI, and SAR using a single radar system. Most evolutionary algorithms, GA's being no exception, include constraints. In practical problems such as waveform design, these constraints are based on physical limitations. By developing objective functions that are mapped to mission performance via waveform parameters (the solution space for our application) we have developed a method for searching a vast solution space to determine Pareto optimal waveform suites that can be used to simultaneously perform multiple radar missions. We have proposed using orthogonal waveforms as a constraint due to the natural separation of such signals in frequency space as well as practical state-of-the art hardware implementation feasibility.